Fault type vectorization method, device, equipment and storage medium

Abstract

The present invention provides a fault type vectorization method, device, equipment and storage medium, the method comprising: obtaining an alarm log sequence, the alarm log sequence comprising alarm logs sorted by time, the alarm logs being associated with fault types; forming a fault type sequence according to the alarm log sequence, the fault type sequence comprising fault types sorted by time; using the fault type sequence as a training sample to train a word skipping model; according to the trained word skipping model, obtaining vectors and weights of each of the fault types, forming a vector table and a weight table of the fault type. The present invention realizes the vectorization of fault types, so that the correlation between faults can be represented according to vectors.

Description

Background Art

[0002] When a complex system has an anomaly, it is often accompanied by multiple faults or alarms. Clarifying the relationship between faults or alarms is a typical task of the operation system. Since vectors have the property of being operable, there are significant advantages in mapping faults to vectors and analyzing faults based on vectors, such as fault correlation and causality. However, due to technical limitations, there has been a lack of effective means to describe faults with vectors. A simple method is to segment the fault description according to "words" and then count the word frequencies to form vectors. This processing method can only mechanically obtain vector representations and cannot embed rich information in the vectors.

[0003] Therefore, how to realize the vectorization of fault types so as to characterize the correlation between faults according to vectors is a technical problem that needs to be solved urgently by those skilled in the art.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present invention, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0005] In view of the problems in the prior art, the purpose of the present invention is to provide a fault type vectorization method, device, equipment and storage medium, which overcomes the difficulties of the prior art and realizes fault type vectorization, so that the correlation between faults can be represented according to vectors.

[0006] An embodiment of the present invention provides a fault type vectorization method, including:

[0007] Acquire an alarm log sequence, wherein the alarm log sequence includes alarm logs sorted by time, and the alarm logs are associated with fault types;

[0008] forming a fault type sequence according to the alarm log sequence, the fault type sequence comprising fault types sorted by time;

[0009] Using the fault type sequence as a training sample to train the skip-word model;

[0010] According to the trained word skipping model, vectors and weights of each fault type are obtained to form a vector table and a weight table of the fault type.

[0011] In some embodiments of the present application, the using the fault type sequence as a training sample to train the word skipping model includes iteratively training the word skipping model, and each iterative training includes:

[0012] Randomly select k consecutive fault types from the fault type sequence, where k is an integer greater than 1;

[0013] Acquire the vectors of the k fault types according to the vector table;

[0014] Sum the vectors of k-1 fault types among the k fault types, and use an activation function to obtain an intermediate vector;

[0015] Using an intermediate vector to perform inner product calculation on the weight table to obtain an intermediate value of each fault type;

[0016] Performing exponential change and probability calculation on the intermediate value of each of the fault types to obtain the probability of another fault type among the k fault types;

[0017] The obtained probability is maximized to update the vector table and the weight table.

[0018] In some embodiments of the present application, the vector table is initialized based on one-hot encoding.

[0019] In some embodiments of the present application, a gradient descent algorithm is used to maximize the obtained probability.

[0020] In some embodiments of the present application, the fault type is obtained based on attribute information of the alarm log.

[0021] In some embodiments of the present application, in the fault type sequence, the more adjacent the fault types are, the closer the distances between the vectors of the fault types are.

[0022] In some embodiments of the present application, after forming the vector table and the weight table of fault types, the following steps are further included:

[0023] Querying the vector table to obtain a vector of a fault type to be processed;

[0024] The fault types to be processed are clustered according to their vector distances.

[0025] According to another aspect of the present application, a fault type vectorization device is also provided, including:

[0026] A first acquisition module is configured to acquire an alarm log sequence, wherein the alarm log sequence includes alarm logs sorted by time, and the alarm logs are associated with fault types;

[0027] A first forming module is configured to form a fault type sequence according to the alarm log sequence, wherein the fault type sequence includes fault types sorted by time;

[0028] A training module, configured to use the fault type sequence as a training sample to train a skip-word model;

[0029] The second forming module is configured to obtain the vectors and weights of each of the fault types according to the trained word skipping model, and form a vector table and a weight table of the fault types.

[0030] According to another aspect of the present invention, there is also provided a fault type vector processing device, comprising:

[0031] processor;

[0032] a memory storing executable instructions of the processor;

[0033] The processor is configured to perform the steps of the fault type vectorization method described above by executing the executable instructions.

[0034] An embodiment of the present invention further provides a computer-readable storage medium for storing a program, wherein the program implements the steps of the above-mentioned fault type vectorization method when executed.

[0035] Compared with the prior art, the purpose of the present invention is:

[0036] The present application forms a fault type sequence based on the alarm log sequence, so that the fault type sequence can be used as a training sample to train the word skipping model, and based on the trained word skipping model, the vector and weight of each fault type are obtained to form a vector table and weight table of the fault type. Since the training sample is a time-based fault type sequence, the vector of the fault type obtained by training can characterize the temporal correlation of the fault type. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Other features, objectives and advantages of the present invention will become more apparent from a reading of the detailed description of non-limiting embodiments made with reference to the following accompanying drawings.

[0038] Figure 1 It is a flow chart of an embodiment of the fault type vectorization method of the present invention.

[0039] Figure 2 It is a flow chart of another embodiment of the fault type vectorization method of the present invention.

[0040] Figure 3 It is a schematic diagram of one iteration of the training word skipping model of the fault type vectorization method of the present invention.

[0041] Figure 4 It is a module diagram of an embodiment of a fault type vectorization device of the present invention.

[0042] Figure 5 It is a module diagram of another embodiment of the fault type vectorization device of the present invention.

[0043] Figure 6 It is a structural schematic diagram of the fault type vectorization device of the present invention.

[0044] Figure 7 It is a schematic diagram of the structure of a computer-readable storage medium according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present invention will be comprehensive and complete and fully convey the concepts of the example embodiments to those skilled in the art. The same reference numerals in the figures represent the same or similar structures, and thus their repeated description will be omitted.

[0046] See also Figure 1 , Figure 1 FIG. 1 is a flow chart of an embodiment of a method for vectorizing fault types of the present invention. An embodiment of the present invention provides a method for vectorizing fault types, comprising the following steps:

[0047] Step S110: Acquire an alarm log sequence, wherein the alarm log sequence includes alarm logs sorted by time, and the alarm logs are associated with fault types.

[0048] Specifically, step S110 can be implemented by collecting system failures, thereby using the alarm logs when the system fails as a data source for the following steps.

[0049] Specifically, in the alarm log sequence, each alarm information can be used as an item of data in the alarm log sequence.

[0050] Specifically, each alarm information may be associated with a fault type. The fault type may be determined, for example, by attributes in the alarm log (such as Interface Down, interface shutdown, etc.).

[0051] Step S120: forming a fault type sequence according to the alarm log sequence, wherein the fault type sequence includes fault types sorted by time.

[0052] Specifically, step S120 may form a time-ordered fault type sequence according to the fault types associated with the alarm log sequence.

[0053] Step S130: using the fault type sequence as a training sample to train the word skipping model.

[0054] Step S140: according to the trained word skipping model, vectors and weights of the fault types are obtained to form a vector table and a weight table of the fault types.

[0055] The present invention forms a fault type sequence based on an alarm log sequence, so that the fault type sequence can be used as a training sample to train a word skipping model, and based on the trained word skipping model, vectors and weights of each fault type are obtained to form a vector table and a weight table of fault types. Since the training sample is a time-based fault type sequence, the vector of the fault type obtained by training can characterize the temporal correlation of the fault type.

[0056] See below Figure 2 , Figure 2 It is a flow chart of another embodiment of the fault type vectorization method of the present invention.

[0057] Step S210: Acquire an alarm log sequence, wherein the alarm log sequence includes alarm logs sorted by time, and the alarm logs are associated with fault types.

[0058] Specifically, step S110 can be implemented by collecting system failures, thereby using the alarm logs when the system fails as a data source for the following steps.

[0059] Specifically, in the alarm log sequence, each alarm information can be used as an item of data in the alarm log sequence.

[0060] Specifically, each alarm information may be associated with a fault type. The fault type may be determined, for example, by attributes in the alarm log (such as Interface Down, interface shutdown, etc.).

[0061] Step S220: forming a fault type sequence according to the alarm log sequence, wherein the fault type sequence includes fault types sorted by time.

[0062] Specifically, step S120 may form a time-ordered fault type sequence according to the fault types associated with the alarm log sequence.

[0063] Step S230: Using the fault type sequence as a training sample, iteratively train the word skipping model. Each iterative training includes the following steps:

[0064] Step S231: randomly selecting k consecutive fault types from the fault type sequence, where k is an integer greater than 1.

[0065] Step S232: obtaining the vectors of the k fault types according to the vector table.

[0066] Specifically, the vector table is initialized based on one-hot encoding. Specifically, one-hot encoding can be used as the initialization code for each fault type, and based on the position where the value of the one-hot encoding is 1, it is used as a vector index and can be mapped to a matrix (vector table). The vector table queried in step S232 is the vector table obtained by the previous iterative update.

[0067] Specifically, the vectors in the vector table may be vectors with the fault type as the central word, and the weights in the weight table may be vectors with the fault type as the background word. The weight table may be initialized as needed.

[0068] Step S233: sum the vectors of k-1 fault types among the k fault types, and use an activation function to obtain an intermediate vector.

[0069] Specifically, the activation function can be selected as needed, for example, the activation function can be a softmax function. This application is not limited to this.

[0070] Step S234: using the intermediate vector to perform inner product calculation on the weight table to obtain the intermediate value of each fault type.

[0071] Specifically, the inner product calculation is performed between the intermediate vector and each fault type in the weight table, so that the intermediate value of each fault type can be obtained.

[0072] Step S235: performing exponential change and probability calculation on the intermediate values ​​of each of the fault types to obtain the probability of another fault type among the k fault types.

[0073] Specifically, exponential change refers to the process of calculating an exponential function with e as the base and each intermediate value as the exponent.

[0074] Specifically, the probability calculation is implemented based on the exponential change of the intermediate value. The probability of each fault type is the ratio of the value of the exponential function of the fault type to the sum of the values ​​of the exponential functions of each fault type.

[0075] Specifically, according to the probability of each fault type, the probability of obtaining another fault type among the k fault types can be calculated.

[0076] Step S236: Maximize the obtained probability to update the vector table and the weight table.

[0077] Specifically, the present application may use a gradient descent algorithm to maximize the obtained probability.

[0078] Step S240: according to the trained word skipping model, the vectors and weights of the fault types are obtained to form a vector table and a weight table of the fault types.

[0079] Step S250: query the vector table to obtain a vector of the fault type to be processed.

[0080] Step S260: clustering the fault types to be processed according to the vector distances of the fault types to be processed.

[0081] Specifically, in the fault type sequence, the more adjacent the fault types are, the closer the distances between the vectors of the fault types are. Therefore, operations such as clustering, compression, and fault analysis of the fault types can be performed based on the vector distances of the fault types.

[0082] Figure 3 It is a schematic diagram of one iteration of the training word skipping model of the fault type vectorization method of the present invention.

[0083] First, randomly select a position in the fault type sequence, and then take k fault types in succession, such as fault X, fault T, fault C, fault D, and then look up the vector table to obtain the corresponding vectors X, T, C, and D. Sum X, C, and T and use the activation function to obtain a new vector Y. Use vector Y to perform the inner product of the weights (weight table) of all fault types to obtain m values ​​(m is the number of fault types), and then perform an exponential transformation on the m values, and then probabilize them to obtain the probability Pd corresponding to fault D. Since the target probability Pd is a function of the vector and the weight, gradient descent can be used to maximize Pd. Therefore, according to the solution of maximizing Pd, the vector table and the weight table can be updated.

[0084] The number of model iteration training can be set as needed. For example, the maximum number of model iterations can be set; or the model iteration training can be stopped when the loss function of the model converges to a set threshold. This application can implement more variations that are not described here.

[0085] Therefore, compared with the prior art, this application performs vector embedding at the fault level to obtain a vector representation for the entire fault, rather than constructing vectors for words (such as words in the fault description) like typical text embedding. In addition, the time-related characteristics are fully exploited to capture the intrinsic semantics.

[0086] The above is only a schematic description of the specific implementation of the present invention, and the present invention is not limited thereto. The splitting, merging, changing of the execution order of steps, splitting, merging of modules, and changing of information transmission are all within the protection scope of the present invention.

[0087] See below Figure 4 , Figure 4 FIG. 4 is a schematic diagram of a module of an embodiment of a fault type vectorization device of the present invention. Figure 4As shown, it includes but is not limited to: a first acquisition module 410, a first formation module 420, a training module 430 and a second formation module 440.

[0088] The first acquisition module 410 is configured to acquire an alarm log sequence, the alarm log sequence comprising alarm logs sorted by time, the alarm logs being associated with a fault type;

[0089] The first forming module 420 is configured to form a fault type sequence according to the alarm log sequence, wherein the fault type sequence includes fault types sorted by time;

[0090] The training module 430 is configured to use the fault type sequence as a training sample to train the skip-word model;

[0091] The second forming module 440 is configured to obtain the vectors and weights of each of the fault types according to the trained word skipping model, and form a vector table and a weight table of the fault types.

[0092] The implementation principle of the above modules can be found in the relevant introduction of the fault type vectorization method, which will not be repeated here.

[0093] The fault type vectorization device of the present invention forms a fault type sequence based on an alarm log sequence, so that the fault type sequence can be used as a training sample to train a word skipping model, and based on the trained word skipping model, the vector and weight of each fault type are obtained to form a vector table and a weight table of the fault type. Since the training sample is a time-based fault type sequence, the vector of the fault type obtained by training can characterize the temporal correlation of the fault type.

[0094] See below Figure 5 , Figure 5 Schematic diagram of another embodiment of the fault type vectorization device of the present invention. The fault type vectorization device 500 of the present invention includes but is not limited to: a first acquisition module 510, a first formation module 520, a training module 530, a second formation module 540, a query module 550 and a clustering module 560. The training module 530 includes but is not limited to: a selection module 531, a second acquisition module 532, an intermediate vector calculation module 533, an intermediate value calculation module 534, a probability calculation module 535 and a maximization module 536.

[0095] The first acquisition module 510 is configured to acquire an alarm log sequence, where the alarm log sequence includes alarm logs sorted by time, and the alarm logs are associated with fault types.

[0096] The first forming module 520 is configured to form a fault type sequence according to the alarm log sequence, where the fault type sequence includes fault types sorted by time.

[0097] The training module 530 is configured to use the fault type sequence as a training sample to iteratively train the word skipping model. The training module 530 includes the following modules to perform each iterative training:

[0098] The selection module 531 is configured to randomly select k consecutive fault types from the fault type sequence, where k is an integer greater than 1.

[0099] The second acquisition module 532 is configured to acquire the vectors of the k fault types according to the vector table.

[0100] The intermediate vector calculation module 533 is configured to sum the vectors of k-1 fault types among the k fault types and obtain an intermediate vector using an activation function.

[0101] The intermediate value calculation module 534 is configured to use the intermediate vector to perform inner product calculation on the weight table to obtain the intermediate value of each of the fault types.

[0102] The probability calculation module 535 is configured to perform exponential change and probability calculation on the intermediate value of each of the fault types to obtain the probability of another fault type among the k fault types.

[0103] The maximization module 536 is configured to maximize the obtained probability to update the vector table and the weight table.

[0104] The second forming module 540 is configured to obtain the vectors and weights of each of the fault types according to the trained word skipping model, and form a vector table and a weight table of the fault types.

[0105] The query module 550 is configured to query the vector table to obtain a vector of a fault type to be processed.

[0106] The clustering module 560 is configured to cluster the fault types to be processed according to the vector distances of the fault types to be processed.

[0107] The implementation principle of the above modules can be found in the relevant introduction of the fault type vectorization method, which will not be repeated here.

[0108] Figure 4 and Figure 5 The fault type vectorization devices 400 and 500 provided by the present invention are only schematically shown. Without violating the concept of the present invention, the splitting, merging, and adding of modules are all within the protection scope of the present invention. The fault type vectorization devices 400 and 500 provided by the present invention can be implemented by software, hardware, firmware, plug-ins, and any combination thereof, and the present invention is not limited thereto.

[0109] The embodiment of the present invention further provides a fault type vectorization processing device, comprising a processor and a memory, wherein executable instructions of the processor are stored, wherein the processor is configured to execute the steps of the fault type vectorization method by executing the executable instructions.

[0110] As shown above, the fault type vectorization processing device of the present invention in this embodiment forms a fault type sequence based on the alarm log sequence, so that the fault type sequence can be used as a training sample to train the word skipping model, and based on the trained word skipping model, the vector and weight of each fault type are obtained to form a vector table and weight table of the fault type. Since the training sample is a time-based fault type sequence, the vector of the fault type obtained by training can characterize the temporal correlation of the fault type.

[0111] It will be appreciated by those skilled in the art that various aspects of the present invention may be implemented as systems, methods or program products. Therefore, various aspects of the present invention may be specifically implemented in the following forms, namely: complete hardware implementation, complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, which may be collectively referred to herein as "circuits", "modules" or "platforms".

[0112] Figure 6 Schematic diagram of the structure of the fault type vector processing device of the present invention. Figure 6 The electronic device 600 according to this embodiment of the present invention is described. Figure 6 The electronic device 600 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0113] like Figure 6 As shown, the electronic device 600 is in the form of a general computing device. The components of the electronic device 600 may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including the storage unit 620 and the processing unit 610), a display unit 640, etc.

[0114] The storage unit stores program codes, which can be executed by the processing unit 610, so that the processing unit 610 executes the steps according to various exemplary embodiments of the present invention described in the above fault type vectorization method section of this specification. For example, the processing unit 610 can execute the following steps: Figure 1 Follow the steps shown in .

[0115] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 6201 and / or a cache memory unit 6202 , and may further include a read-only memory unit (ROM) 6203 .

[0116] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0117] Bus 630 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0118] The electronic device 600 may also communicate with one or more external devices 6001 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 650. Furthermore, the electronic device 600 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 via a bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms, etc.

[0119] The embodiment of the present invention further provides a computer-readable storage medium for storing a program, and the steps of the fault type vectorization method are implemented when the program is executed. In some possible implementations, various aspects of the present invention can also be implemented in the form of a program product, which includes a program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps of various exemplary embodiments of the present invention described in the above fault type vectorization method section of this specification.

[0120] As shown above, the computer-readable storage medium for performing fault type vectorization in this embodiment forms a fault type sequence based on the alarm log sequence, so that the fault type sequence can be used as a training sample to train the word skipping model, and based on the trained word skipping model, the vector and weight of each fault type are obtained to form a vector table and a weight table of the fault type. Since the training sample is a time-based fault type sequence, the vector of the fault type obtained by training can characterize the temporal correlation of the fault type.

[0121] Figure 7 Schematic diagram of the structure of the computer-readable storage medium of the present invention. Figure 7 As shown, a program product 700 for implementing the above method according to an embodiment of the present invention is described, which can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited thereto. In this document, a readable storage medium can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, an apparatus or a device.

[0122] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0123] Computer readable storage media may include data signals propagated in baseband or as part of a carrier wave, wherein readable program codes are carried. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or device. The program codes contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof.

[0124] Program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0125] In summary, the present application forms a fault type sequence based on the alarm log sequence, so that the fault type sequence can be used as a training sample to train the word skipping model, and based on the trained word skipping model, the vector and weight of each fault type are obtained to form a vector table and weight table of the fault type. Since the training sample is a time-based fault type sequence, the vector of the fault type obtained by training can characterize the temporal correlation of the fault type.

[0126] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A fault type vectorization method, characterized in that: include: Acquire an alarm log sequence, wherein the alarm log sequence includes alarm logs sorted by time, and the alarm logs are associated with fault types; forming a fault type sequence according to the alarm log sequence, the fault type sequence comprising fault types sorted by time; Using the fault type sequence as a training sample to train the skip-word model; According to the trained skip-word model, vectors and weights of each fault type are obtained to form a vector table and a weight table of the fault type; The method of using the fault type sequence as a training sample to train the word skipping model includes iterative training of the word skipping model, wherein each iterative training includes: Randomly select k consecutive fault types from the fault type sequence, where k is an integer greater than 1; Acquire the vectors of the k fault types according to the vector table; Sum the vectors of k-1 fault types among the k fault types, and use an activation function to obtain an intermediate vector; Using an intermediate vector to perform inner product calculation on the weight table to obtain an intermediate value of each fault type; Performing exponential change and probability calculation on the intermediate value of each of the fault types to obtain the probability of another fault type among the k fault types; The obtained probability is maximized to update the vector table and the weight table.

2. The fault type vectorization method according to claim 1, characterized in that: The vector table is initialized based on one-hot encoding.

3. The fault type vectorization method according to claim 1, characterized in that: The gradient descent algorithm is used to maximize the obtained probability.

4. The fault type vectorization method according to claim 1, characterized in that: The fault type is obtained based on the attribute information of the alarm log.

5. The fault type vectorization method according to any one of claims 1 to 4, characterized in that: In the fault type sequence, the more adjacent the fault types are, the closer the distances between the vectors of the fault types are.

6. The fault type vectorization method according to any one of claims 1 to 4, characterized in that: After forming the fault type vector table and weight table, it also includes: Querying the vector table to obtain a vector of a fault type to be processed; The fault types to be processed are clustered according to their vector distances.

7. A fault type vectorization device, characterized in that: include: A first acquisition module is configured to acquire an alarm log sequence, wherein the alarm log sequence includes alarm logs sorted by time, and the alarm logs are associated with fault types; A first forming module is configured to form a fault type sequence according to the alarm log sequence, wherein the fault type sequence includes fault types sorted by time; A training module, configured to use the fault type sequence as a training sample to train a skip-word model; A second forming module is configured to obtain the vector and weight of each of the fault types according to the trained word skipping model, and form a vector table and a weight table of the fault types; The method of using the fault type sequence as a training sample to train the word skipping model includes iterative training of the word skipping model, wherein each iterative training includes: Randomly select k consecutive fault types from the fault type sequence, where k is an integer greater than 1; Acquire the vectors of the k fault types according to the vector table; Sum the vectors of k-1 fault types among the k fault types, and use an activation function to obtain an intermediate vector; Using an intermediate vector to perform inner product calculation on the weight table to obtain an intermediate value of each fault type; Performing exponential change and probability calculation on the intermediate value of each of the fault types to obtain the probability of another fault type among the k fault types; The obtained probability is maximized to update the vector table and the weight table.

8. A fault type vector processing device, characterized in that: include: processor; a memory storing executable instructions of the processor; The processor is configured to execute the steps of the fault type vectorization method according to any one of claims 1 to 6 by executing the executable instructions.

9. A computer-readable storage medium for storing a program, characterized in that: When the program is executed, the steps of the fault type vectorization method according to any one of claims 1 to 6 are implemented.

Images